50 WILLIAMS— DEEP KANSAN PONDINGS IN 



which produces beach-lines — both filled and undercut — of good defi- 

 nition, high terraces with long and flat tops, and all the signs of long 

 flooding : the latter will rarely persist at an exact level. We shall 

 come upon good examples in this discussion, and when ice-dams 

 occur in a valley like those of the Juniata or of the Allegheny, where 

 there is a fall of the regional surface commensurate with that of the 

 floodplain of the stream, it is evident that the height of the ice-dams 

 would be similarly influenced, and the sporadic bars and terraces 

 formed in the pondings would have a proportional fall in elevation, 

 and of so marked a character that it would simulate the slope of a 

 high gradation plain. 



The illustrative material to be examined is found in the Hudson, 

 Lehigh and Juniata valleys, whose streams were never reversed, and 

 whose pondings were against ice-dams : in the Bald Eagle, and in an 

 inconsequential part of the West Branch of Susquehanna River, 

 between Williamsport and Lock Haven, both of which were tem- 

 porarily 'reversed, and whose ponding was between the glacier and 

 a col. 



The Allegiheny is a patchwork stream. It flows through 4 val- 

 leys in reverse, and through 4 trenched cols. There were many 

 pondings with surfaces above 2,100 in the Pennsylvania Highlands; 

 above 1,600; above 1.500; above 1,430, and above 1,200, as we pro- 

 ceed from those highlands to the mouth of Clarion River, where the 

 crests of the region are 1,000 feet lower. The terraces, bars, sporadic 

 areas of gravel of Kansan age partake of the elevation of the pond- 

 ing, and fall in elevation, as will be seen below. 



It is generally acknowledged that there was no sinking of the 

 Kansan border in Pennsylvania during glacial times, and that the 

 isobase of 200 feet crosses the Hudson Valley near Storm King 

 Mountain. We can therefore use the Government and State Topo- 

 graphic Quadrangles to measure relative elevations along that 

 border, and, with proper corrections, elsewhere, during the period 

 treated in this paper. The most of the illustrations which follow 

 have been published before. The photographs were taken by the 

 writer between September, 1892, and August, 1897. It remains to 

 acknowledge his great indebtedness to the late Dr. Joseph Barrell 

 who, as an assistant traversed the entire Kansan Border between the 



